The ABINIT Users Mailing List ( CLOSED )

["D. R. Hamann" <drh@bell-labs.com>]

Dear occupierliu (?):

I'm not sure why your acoustic phonons are not closer to zero at gamma, 
but the mathematics used to compute the atom-relaxation contribution to 
the elastic constants projects the 3 acoustic modes out of the 
interatomic-force-constant matrix prior to inverting it, so your 
relaxed-atom elastic constants should not be sensitive to the explicit 
errors this warning flags.  Since you've done convergence testing 
already, simply look at the elastic constants.  If they appear to be 
converged, they are.  The warning about the acoustic mode eigenvalues 
serves as a possible indication that the phonons that matter for the 
relaxation correction, the various optic modes, might not be converged 
either, so at least some testing is warranted when the warning appears.

I hope this helps, and that your elastic constants are adequately converged.

Don Hamann

occupierliu@yahoo.com.cn wrote:

> Dear all:
>    I'm trying to calculate the elastic constants of PbTiO3(P4mm) and in the 
> output file of anaddb there is a warning message:
> phonon modes warning- :
> accoustic sum rule violation met:the eigenvalues of accoustic mode
> are too large at Gamma point
> increase cutoff energy or k-points sampling.
> the three eigenvalues are:    2.480398E-04    3.065292E-05    2.480407E-04.
>
> So I increased the ecut from 45H to 55H and ngkpt from 75 to 115 while the 
> total energe decrease from -1.65976000880566E+02H to -1.6598600852E+02H and 
> -1.6597872349E+02, respectively. I think the convergence is good enough, but 
> unfortunately the warning message is still there only the the eigenvalues 
> change a litte:
> ecut 45 and ngkpt 115    2.480462E-04    3.065251E-05    2.480464E-04
> ecut 55 and ngkpt 75     1.718815E-04    3.278633E-06    1.718815E-04.
>
> Should I keep increaseing the ecut and ngkpt untill this warning disappears 
> or there this some other parameters I can change? Because the large ecut and 
> ngkpt cause a huge time-consuming computation work.
>    Thanks a lot in advance.
>
> # Here My input file for ABINIT 4.6.4
> # PbTiO3 (No.99, P4mm)
>
>   ndtset     2
>
> #Dataset1: Initial self-consistent run
>    iscf1     5
>  kptopt1     1
>  tolvrs1     1.0d-18
>
> #Dataset2: Response Function calculation
>  getwfk2    -1
>    iscf2     5
>  kptopt2     2
>    nqpt2     1
>     qpt2     0 0 0 
>  rfphon2     1
> rfatpol2     1 5
>  rfstrs2     3
>   rfdir2     1 1 1
>  tolvrs2     1.0d-12
>
> #SCF calculation
>    ngkpt     10 10 10
>    nband     22
>
> #Definition of the planewave basis set
>     ecut     45.0
>   ecutsm     0.5
>
> #Definition of the SCF procedure
>    nstep     100
>   tolvrs     1.0d-18 
>   diemac     6.0 
>    
> #Common Data Definition of the unit cell
>    acell     7.3016242907  7.3016242907  7.5097985279
>    rprim     1.0     0.0     0.0
>              0.0     1.0     0.0
>              0.0     0.0     1.0
>   ntypat     3
>    znucl     82 8 22
>    natom     5
>    typat     1 3 2 2 2
>     xred     0.0000000000E+00  0.0000000000E+00  0.0000000000E+00
>              5.0000000000E-01  5.0000000000E-01  4.6901801570E-01
>              5.0000000000E-01  5.0000000000E-01  9.2453581448E-01
>              5.0000000000E-01  8.7581154020E-47  4.0875625165E-01
>              0.0000000000E+00  5.0000000000E-01  4.0875625165E-01
>  


--
D. R. Hamann                            Phone: 908-582-4454
Director, Theoretical Materials         Fax:   908-582-4868
 Physics Research (retired)             email: drh@bell-labs.com
Bell Laboratories 
Lucent Technologies
700 Mountain Ave, Room 1D-371
Murray Hill, NJ 07974-0636 USA